36 GROTE — SPECIALIZATIONS OF LEPIDOPTEROUS WING. [April?, 



vein 'Mx" of primaries well developed. This vein arises at base 

 next to vii, runs outwardly and downwardly, in a more or less 

 strongly given cur/e, to internal margin. I am inclined to regard 

 this vein ''ix" as the remnant of a longitudinal vein which has 

 become shortened, in the same way as vii becomes shortened and 

 bent on the hind wings of the Papilionides, and of which latter the 

 stages between the long, straight vein and the shorter curved vein 

 are extant. This vein ''ix" is opposed in position to vein viii of 

 the Hesperiades, which appears as a slender fork to vein vii at base 

 and is more or less distinct in the Pieridae, Limnadidse, Heli- 

 coniidae, Libytheidae, Nemeobiidse, Riodinidae, Lycsenidae and 

 Hesperiad^. I have found traces of it in Argynnis and Vanessa 

 among the Nymphalidae proper, but no indication of it in the 

 Agapetidae. Where it is wanting in otherwise related forms I 

 adopt the view that it has faded out. It has faded out, then, in 

 J^orpho, while the cubital cross-vein is persistent. 



While vein '' ix " appears to be a character of a primary nature, 

 the hind wings of the Papilionides display two other features, 

 which, since they recur in other butterflies or moths, may be con- 

 sidered as characters of convergence. These are the humeral cell 

 and the single internal vein. The three neurational features of the 

 Papilionides together are wanting in all other butterflies, but the 

 single internal vein appears again in the Saturniades, the humeral 

 cell in the Limnadidae and Bombycides, while nowhere throughout 

 the higher Lepidoptera, so far as I yet know, is the primary char- 

 acter of vein " ix " repeated. The ancestry of the Papilionides is 

 not yet made out by the discovery of surviving forms. 



It is impossible to leave the Papilionides in a classificatory posi- 

 tion between the Blues and the Skippers, because such a position 

 violates the integrity of the Lyc^nid-Hesperian branch. The 

 diurnals, as a whole, are susceptible of a tripartite division upon 

 the neurational characteristics. These three divisions are, severally : 

 the Parnassi-Papilionidae, the Pieri-Nymphalidce, and the Lycasni- 

 Hesperiadae. The first of these is isolated by the presence of vein 

 *' ix," the last two have a common bond in the presence of vein 

 viii of the fore wings. Therefore the value of the divisions is 

 unequal, and the first outweighs either of the two latter. If the 

 Hesperiades are, then, to be forced at any point to admit the 

 Papilionides, that point might be selected between the Pieri- 

 Nymphalid and Lycaeni-Hesperid groups. But the principal 



